Pre-inquiry method and device, storage medium and equipment

Abstract

The invention discloses a pre-inquiry method and device, a storage medium and equipment. According to the scheme, on the basis of description information input by a to-be-inquired user, symptom information of the to-be-inquired user is determined through a large language model; extracting the symptom information of the to-be-inquired user by using a natural language processing technology, determining a symptom attribute problem corresponding to an extraction processing result according to the extraction processing result, and constructing a problem queue based on the symptom attribute problem; and determining a pre-inquiry result of the to-be-inquired user based on the extraction processing result and the question queue. In the prior art, pre-inquiry is realized by completely depending on a large language model, the training process of the large language model is complicated, and parameters are fixed after training. Compared with the prior art that the pre-inquiry result is not accurate, the method has obvious advantages.

Description

Technical Field

[0001] The present application relates to the field of intelligent medical technology, and in particular to a pre-diagnosis method, device, storage medium and equipment. Background Art

[0002] With the continuous development of medical technology and the intensification of population aging, medical resources are becoming increasingly tight. At present, hospital outpatient clinics generally face problems such as low efficiency in patient information collection, cumbersome medical record writing, long patient waiting time, and high work pressure on doctors. In order to improve diagnosis and treatment efficiency, optimize patient experience, reduce medical costs, promote doctor-patient communication, and improve medical quality, a pre-consultation method is proposed, namely the pre-consultation method, which aims to achieve initial communication between doctors and patients through online communication and improve the efficiency and quality of medical services. At the same time, pre-consultation can also help doctors understand the patient's condition more comprehensively and provide a basis for formulating scientific and reasonable treatment plans. With the continuous advancement of technology and the in-depth advancement of medical reform, pre-consultation will play a more important role in the future.

[0003] Online questionnaires are a common form of pre-consultation. Patients can answer the corresponding questions by clicking on the answers on the mobile app or app before seeing the doctor. However, this method has the defects of relatively fixed questions and no specificity, and cannot meet the doctor's requirements for prior understanding of certain diseases. Later, a pre-consultation method based on knowledge graphs and large language models appeared. This method generates corpus by constructing knowledge graphs, and uses the corpus to fine-tune the large language model. Using the fine-tuned large language model as the basis, the large language model automatically outputs questions and adjusts subsequent questions according to the patient's answers for pre-consultation. However, this method also has defects: on the one hand, the quality of the questions asked depends heavily on the data set and the large language model. In order to achieve good results, a large amount of fine-tuning corpus needs to be constructed, and the process is complicated and incomplete; on the other hand, the list of questions cannot be customized, because specific attribute questions need to be asked for specific diseases, and the parameters of the fine-tuned large model are fixed.

[0004] How to improve the accuracy of pre-diagnosis results is a technical problem that needs to be solved urgently. Summary of the invention

[0005] Based on the above problems, the present application provides a pre-diagnosis method, apparatus, storage medium and device, the purpose of which is to improve the accuracy of pre-diagnosis results.

[0006] The embodiments of the present application disclose the following technical solutions:

[0007] The first aspect of the present application provides a pre-diagnosis method, the method comprising:

[0008] Based on the description information input by the user to be interviewed, use a large language model to determine the symptom information of the user to be interviewed;

[0009] Use natural language processing technology to extract and process the symptom information of the user to be interviewed, determine the symptom attribute questions corresponding to the extraction and processing results, and construct a question queue based on the symptom attribute questions;

[0010] Determine the preliminary interview result of the user to be interviewed based on the extraction and processing results and the question queue.

[0011] Optionally, determining the preliminary interview result of the user to be interviewed based on the extraction and processing results and the question queue includes:

[0012] Judge whether there is a question corresponding to the extraction and processing result in the question queue to obtain a first judgment result;

[0013] If the first judgment result is yes, remove the question corresponding to the extraction and processing result in the question queue to obtain the question queue after removal, and determine the results corresponding to each question in the question queue after removal;

[0014] Summarize the results corresponding to each question in the obtained question queue after removal according to a preset rule, and output the preliminary interview result;

[0015] If the first judgment result is no, determine the results corresponding to each question in the question queue, and summarize the results corresponding to each obtained question according to a preset rule, and output the preliminary interview result.

[0016] Optionally, based on the description information input by the user to be interviewed, using a large language model to determine the symptom information of the user to be interviewed includes:

[0017] Use a large language model to judge whether the description information is related to preset interview questions to obtain a second judgment result;

[0018] If the second judgment result is yes, use the large language model to convert the description information to obtain the symptom information of the user to be interviewed;

[0019] If the second judgment result is no, the user to be interviewed re-enters the description information.

[0020] Optionally, using natural language processing technology to extract and process the symptom information of the user to be interviewed, determining the symptom attribute questions corresponding to the extraction and processing results, and constructing a question queue based on the symptom attribute questions includes:

[0021] Perform word segmentation on the symptom information of the user to be interviewed to obtain the word segmentation result;

[0022] Perform information extraction on the word segmentation result to obtain the information extraction result;

[0023] Determine the symptom attribute questions corresponding to the information extraction result from the knowledge base according to the information extraction result, and construct a question queue based on the symptom attribute questions; the knowledge base includes symptom attribute questions of various symptoms.

[0024] Optionally, before determining the symptom information of the user to be interviewed using a large language model based on the description information input by the user to be interviewed, it further includes:

[0025] Obtain the basic information of the user to be interviewed; the pre-interview result corresponds to the basic information.

[0026] Optionally, before determining the symptom information of the user to be interviewed using a large language model based on the description information input by the user to be interviewed, it further includes:

[0027] Judge whether the user to be interviewed uploads a picture to obtain a third judgment result;

[0028] If the third judgment result is yes, use OCR technology to recognize the picture to obtain the recognition result, and upload the recognition result to the target system;

[0029] If the third judgment result is no, determine the symptom information of the user to be interviewed using a large language model based on the description information input by the user to be interviewed.

[0030] The second aspect of this application provides a pre-interview device, which includes:

[0031] A symptom information determination module, configured to determine the symptom information of the user to be interviewed using a large language model based on the description information input by the user to be interviewed;

[0032] A question queue construction module, configured to perform extraction processing on the symptom information of the user to be interviewed using natural language processing technology, determine the symptom attribute questions corresponding to the extraction processing result, and construct a question queue based on the symptom attribute questions;

[0033] A pre-interview result determination module, configured to determine the pre-interview result of the user to be interviewed based on the extraction processing result and the question queue.

[0034] Optionally, the pre-interview result determination module is specifically configured to:

[0035] Judge whether there is a question corresponding to the extraction processing result in the question queue to obtain a first judgment result;

[0036] If the first judgment result is yes, remove the question corresponding to the extraction processing result from the question queue to obtain the question queue after removal, and determine the result corresponding to each question in the question queue after removal;

[0037] Output the result corresponding to each question in the obtained question queue after removal as the pre-consultation result;

[0038] If the first judgment result is no, determine the result corresponding to each question in the question queue, and output the result as the pre-consultation result.

[0039] The third aspect of the present application provides a computer-readable storage medium, in which a computer program is stored. When the program is run by a computer device, the pre-consultation method provided by any implementation manner of the first aspect is implemented.

[0040] The fourth aspect of the present application provides a computer device, which is used to run a computer program. When the program runs, the pre-consultation method provided by any implementation manner of the first aspect is executed.

[0041] Compared with the prior art, the present application has the following beneficial effects:

[0042] The pre-consultation method provided by the present application, based on the description information input by the user to be consulted, uses a large language model to determine the symptom information of the user to be consulted, and uses natural language processing technology to extract and process the symptom information of the user to be consulted. By combining the large language model with natural language processing technology, effective information related to the consultation of the user to be consulted is extracted. According to the extraction processing result, determine the symptom attribute question corresponding to the extraction processing result, and construct a question queue based on the symptom attribute question. In this way, a question queue is formulated for the user to be consulted, so that the questions included in the question queue have strong pertinence. Determine the pre-consultation result of the user to be consulted based on the extraction processing result and the question queue, improving the accuracy of the pre-consultation result. Description of the Drawings

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0044] Figure 1 It is a flowchart of a pre-consultation method provided by an embodiment of the present application;

[0045] Figure 2 It is a flowchart of another pre-consultation method provided by an embodiment of the present application;

[0046] Figure 3 It is a schematic structural diagram of a pre-consultation device provided by an embodiment of the present application;

[0047] Figure 4 It is a schematic structural diagram of a computer device provided by an embodiment of the present application. Detailed implementation manners

[0048] As described above, for the pre-consultation method based on the knowledge graph and the large language model, this method generates corpus by constructing a knowledge graph and uses the corpus to fine-tune the large language model. Using the fine-tuned large language model as the basis, the large language model automatically outputs questions and adjusts subsequent questions according to the patient's answers for pre-consultation. However, this method also has defects: on the one hand, the quality of the consultation questions depends heavily on the data set and the large language model. To achieve good results, a large amount of fine-tuning corpus needs to be generated, and the process is complex and incomplete; on the other hand, the question list cannot be customized because specific attribute questions need to be asked for specific diseases, and the parameters of the fine-tuned large model are fixed.

[0049] In view of the above problems, through research, the inventor has proposed a pre-consultation method, device, storage medium and equipment. Based on the description information input by the user to be consulted, the symptom information of the user to be consulted is determined by using a large language model; the symptom information of the user to be consulted is extracted and processed by using natural language processing technology, and symptom attribute questions corresponding to the extraction and processing results are determined according to the extraction and processing results, and a question queue is constructed based on the symptom attribute questions; the pre-consultation result of the user to be consulted is determined based on the extraction and processing results and the question queue.

[0050] In order to enable those skilled in the art to better understand the solution of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0051] See Figure 1 , this figure is a flowchart of a pre-consultation method provided by an embodiment of the present application. As Figure 1 shown, the method includes the following steps:

[0052] S101. Based on the description information input by the user to be consulted, use a large language model to determine the symptom information of the user to be consulted.

[0053] Among them, based on the description information input by the user to be questioned, the large language model is used to determine the symptom information of the user to be questioned, including:

[0054] Use the large language model to determine whether the description information is relevant to the preset interrogation questions, and obtain a second judgment result;

[0055] If the second judgment result is yes, use the large language model to convert the description information to obtain the symptom information of the user to be questioned;

[0056] If the second judgment result is no, the user to be questioned re-enters the description information.

[0057] The user to be questioned can describe by means of clicking, direct input, etc. Use the large model to determine whether the description information input by the user to be questioned is relevant to the preset interrogation questions. For example, the preset interrogation question is "What symptoms or discomfort do you have?" The description information input by the user to be questioned is "I have a headache for three days and it hasn't gotten better." Use the large language model to determine that the description information is relevant to the preset interrogation question, and use the large language model to convert the description information to obtain the symptom information of the user to be questioned as "The patient has had a headache for three days without relief." If it is not relevant, the user to be questioned re-enters the description information.

[0058] S102. Use natural language processing technology to extract and process the symptom information of the user to be questioned, determine the symptom attribute questions corresponding to the extraction and processing results according to the extraction and processing results, and construct a question queue based on the symptom attribute questions.

[0059] Among them, using natural language processing technology to extract and process the symptom information of the user to be questioned, determining the symptom attribute questions corresponding to the extraction and processing results according to the extraction and processing results, and constructing a question queue based on the symptom attribute questions, including:

[0060] Perform word segmentation processing on the symptom information of the user to be questioned to obtain a word segmentation processing result;

[0061] Perform information extraction on the word segmentation processing result to obtain an information extraction result;

[0062] Determine the symptom attribute questions corresponding to the information extraction result from the knowledge base according to the information extraction result, and construct a question queue based on the symptom attribute questions; the knowledge base includes symptom attribute questions of various symptoms.

[0063] Word segmentation is to break the input text into basic language units (such as words or phrases) for subsequent analysis and processing. Information extraction is performed on the result of word segmentation, aiming to identify key symptoms and related attributes of the symptoms (such as the type of symptom, duration, severity, etc.). Based on the extracted results, the symptom attribute problems corresponding to these symptoms are determined from a pre-established knowledge base, which includes symptom attribute problems of various symptoms. For example, retrieving the attribute problems related to "headache" from the knowledge base includes the location of symptom occurrence, the nature of the symptom, and the frequency of the symptom, etc. According to the determined symptom attribute problems, a question queue is constructed. This question queue will serve as the basis for further questioning by doctors later to help understand the patient's condition more comprehensively.

[0064] Using natural language processing technology to extract and process the symptom information of the user to be questioned, determining the symptom attribute problems corresponding to the extraction and processing results based on the extraction and processing results, and constructing a question queue based on the symptom attribute problems. In this way, not only can the key information in the patient's description be effectively extracted, but also a targeted question list can be generated, improving the accuracy and efficiency of medical consultation.

[0065] S103. Determine the preliminary consultation result of the user to be questioned based on the extraction and processing result and the question queue.

[0066] Among them, determining the preliminary consultation result of the user to be questioned based on the extraction and processing result and the question queue includes:

[0067] Judge whether there is a question corresponding to the extraction and processing result in the question queue to obtain a first judgment result;

[0068] If the first judgment result is yes, remove the question corresponding to the extraction and processing result in the question queue to obtain the question queue after removal, and determine the result corresponding to each question in the question queue after removal;

[0069] Summarize the results corresponding to each question in the obtained question queue after removal according to a preset rule, and output the preliminary consultation result;

[0070] If the first judgment result is no, determine the result corresponding to each question in the question queue, and summarize the results corresponding to each obtained question according to a preset rule, and output the preliminary consultation result.

[0071] Check the existing question queue, for example, "Is there a fever?", "Is there a cough?", etc., to see if the answers to these questions have been extracted from the symptom information of the user to be interviewed. If so, for example, the user to be interviewed mentions having a fever, then remove this question from the queue. For the remaining unanswered questions, such as "Is there a cough?", continue to ask the user to be interviewed until all questions in the question queue have answers. Then, based on the user's answers to the questions in the question queue, output the pre-interview results, which include the main symptoms, possible disease speculation, and recommended next steps (such as medical advice). This approach ensures that the pre-interview process is both efficient and accurate, can make full use of the existing information of the user to be interviewed, reduce unnecessary repeated questions, and improve the user experience.

[0072] The pre-interview method provided by the embodiment of the present application, based on the description information input by the user to be interviewed, uses a large language model to determine the symptom information of the user to be interviewed, and uses natural language processing technology to extract and process the symptom information of the user to be interviewed. By combining the large language model with natural language processing technology, effective information related to the interview of the user to be interviewed is extracted. According to the extraction and processing results, symptom attribute questions corresponding to the extraction and processing results are determined, and a question queue is constructed based on the symptom attribute questions. In this way, a question queue is formulated for the user to be interviewed, so that the questions included in the question queue are highly targeted. Based on the extraction and processing results and the question queue, the pre-interview results of the user to be interviewed are determined, improving the accuracy of the pre-interview results.

[0073] See Figure 2 , which is a flowchart of another pre-interview method provided by the embodiment of the present application. On the basis of the previous embodiment, the pre-interview method is improved by adding the step of obtaining the basic information of the user to be interviewed. As Figure 2 shown, the method includes the following steps:

[0074] S201. Obtain the basic information of the user to be interviewed.

[0075] The pre-interview results correspond to the basic information.

[0076] Obtain the basic information such as the pre-interview time, department, gender, age, etc. of the user to be interviewed by means of point selection or free input. In this way, the expression needs of the user to be interviewed are met, providing basic information for the subsequent pre-interview process.

[0077] S202. Based on the description information input by the user to be interviewed, use a large language model to determine the symptom information of the user to be interviewed.

[0078] Before determining the symptom information of the user to be interviewed based on the description information input by the user to be interviewed using a large language model, it further includes:

[0079] Judge whether the user to be interviewed uploads a picture to obtain a third judgment result;

[0080] If the third judgment result is yes, use OCR technology to recognize the picture to obtain a recognition result, and upload the recognition result to the target system;

[0081] If the third judgment result is no, determine the symptom information of the user to be interviewed based on the description information input by the user to be interviewed using a large language model.

[0082] Optical Character Recognition (OCR) technology is a technology that can convert the text in various types of documents (such as printed text, handwriting, etc.) into digital text that can be processed by a computer. This technology greatly improves the efficiency of information entry, reduces manual input errors, and enables the information in paper documents to be digitally stored and managed. Use OCR technology to recognize the picture to obtain a recognition result, and upload the recognition result to the target system, where the target system includes an electronic medical record system and a Hospital Information System (HIS), etc. This method allows the user to be interviewed to upload historical medical record content and use OCR technology for recognition, saving the process of the user to be interviewed from describing, and facilitating doctors to understand the historical medical record information of the user to be interviewed.

[0083] Among them, determining the symptom information of the user to be interviewed based on the description information input by the user to be interviewed using a large language model includes:

[0084] Use a large language model to judge whether the description information is relevant to the preset interview questions to obtain a second judgment result;

[0085] If the second judgment result is yes, use the large language model to convert the description information to obtain the symptom information of the user to be interviewed;

[0086] If the second judgment result is no, the user to be interviewed re-enters the description information.

[0087] Although the pre-interview method provided by the embodiments of this application also uses a large language model, it uses a simple corpus to judge the relevance between the preset interview questions and the description information input by the user to be interviewed. Even if the large language model needs to be fine-tuned in the subsequent process, it can be completed in a short time and will not have too much impact on the entire pre-interview process.

[0088] S203. Use natural language processing technology to extract and process the symptom information of the user to be interviewed, determine the symptom attribute questions corresponding to the extraction and processing results according to the extraction and processing results, and construct a question queue based on the symptom attribute questions.

[0089] Among them, using natural language processing technology to extract and process the symptom information of the user to be interviewed, determining the symptom attribute questions corresponding to the extraction and processing results according to the extraction and processing results, and constructing a question queue based on the symptom attribute questions includes:

[0090] Perform word segmentation processing on the symptom information of the user to be interviewed to obtain the word segmentation processing result;

[0091] Perform information extraction on the word segmentation processing result to obtain the information extraction result;

[0092] Determine the symptom attribute questions corresponding to the extraction and processing results from the knowledge base according to the information extraction result, and construct a question queue based on the symptom attribute questions; the knowledge base includes symptom attribute questions of various symptoms.

[0093] For example, the preset interview question is "What symptoms or discomfort do you have?", and the description information input by the user to be interviewed is "I have a headache for three days and it hasn't gotten better". Use the large language model to judge whether the above preset interview question is relevant to the description information input by the user to be interviewed, and the judgment result is yes. Next, use the large language model to convert the description information to obtain the symptom information of "The patient has had a headache for three days without relief". Perform word segmentation processing on the symptom information of the user to be interviewed, and the word segmentation processing result is: patient, headache, three days, no relief; perform information extraction on the processing result to obtain the information extraction result as {... "symptom name": "headache", "whether it has improved or relieved": "no", "symptom duration": "three days"...}; Determine the symptom attribute questions corresponding to headache from the knowledge base according to the information extraction result. For example, the symptom occurrence locations include the top, forehead, temporal region, occipital region, and the whole head, and the location is not fixed. The nature of the symptom is like stabbing pain, dull pain, squeezing pain, etc. Construct a question queue based on the symptom attribute questions, and the question queue is shown in Table 1.

[0094] Table 1

[0095] Please describe any symptoms or discomfort How long has the headache been present? Where does the headache occur (frontal, temporal, parietal, occipital)? What does the headache feel like (e.g., radiating)? Is the headache continuous, intermittent, or occasional? How long does the headache last on average each time? What is the severity of the headache?

[0096] The above consultation questions can be flexibly customized based on the requirements of doctors and departments. It is also convenient to modify the attribute questions related to symptoms in the knowledge base. Once a department needs to ask specific questions about specific symptoms, the pre-consultation method provided in this embodiment can be quickly implemented by modifying the question queue or modifying and adding the attribute questions corresponding to the symptoms in the knowledge base, and it will not affect the entire consultation process.

[0097] S204. Determine the pre-consultation result of the user to be consulted based on the extraction and processing result and the question queue.

[0098] Among them, determining the pre-consultation result of the user to be consulted based on the extraction and processing result and the question queue includes:

[0099] Judge whether there is a question corresponding to the extraction and processing result in the question queue to obtain a first judgment result;

[0100] If the first judgment result is yes, remove the question corresponding to the extraction and processing result in the question queue to obtain the question queue after removal, and determine the result corresponding to each question in the question queue after removal;

[0101] Summarize the results corresponding to each question in the obtained question queue after removal according to a preset rule, and output the pre-consultation result;

[0102] If the first judgment result is no, determine the result corresponding to each question in the question queue, and summarize the results corresponding to each obtained question according to a preset rule, and output the pre-consultation result.

[0103] For example, if the extraction and processing result in the previous step includes "symptom duration": "three days", then remove the second question "How long has the headache occurred?" in the question queue shown in Table 1. This method solves the problem of a large amount of descriptive information freely input by the user to be consulted by performing word segmentation and information extraction on the content standardized by the large language model, and further screens the questions included in the question queue according to the results of natural language technologies such as word segmentation and information extraction, making the entire pre-consultation process simpler and more efficient.

[0104] The pre-consultation method provided by the embodiments of the present application first obtains the basic information of the user to be consulted. For example, the basic information of the user to be consulted is obtained by means of point selection or free input, so as to meet the expression needs of the user to be consulted. Then, based on the description information input by the user to be consulted, the large language model is used to determine the symptom information of the user to be consulted, and the natural language processing technology is used to extract and process the symptom information of the user to be consulted. By combining the large language model with the natural language processing technology, the effective information related to the consultation of the user to be consulted is extracted. According to the extraction and processing results, the symptom attribute questions corresponding to the extraction and processing results are determined, and a question queue is constructed based on the symptom attribute questions. In this way, a question queue is formulated for the user to be consulted, so that the questions included in the question queue are highly targeted. Based on the extraction and processing results and the question queue, the pre-consultation result of the user to be consulted is determined, improving the accuracy of the pre-consultation result.

[0105] Based on the pre-consultation method introduced in the foregoing embodiments, correspondingly, the present application also provides a pre-consultation device. Figure 3 As shown in the structure diagram of the device. Figure 3 As shown, the pre-consultation device includes:

[0106] A symptom information determination module 301, configured to determine the symptom information of the user to be consulted by using a large language model based on the description information input by the user to be consulted;

[0107] A question queue construction module 302, configured to extract and process the symptom information of the user to be consulted by using natural language processing technology, determine the symptom attribute questions corresponding to the extraction and processing results according to the extraction and processing results, and construct a question queue based on the symptom attribute questions;

[0108] A pre-consultation result determination module 303, configured to determine the pre-consultation result of the user to be consulted based on the extraction and processing results and the question queue.

[0109] Optionally, the pre-consultation result determination module is specifically configured to:

[0110] Judge whether there is a question corresponding to the extraction and processing result in the question queue, and obtain a first judgment result;

[0111] If the first judgment result is yes, remove the question corresponding to the extraction and processing result in the question queue to obtain the question queue after removal, and determine the result corresponding to each question in the question queue after removal;

[0112] Output the result corresponding to each question in the obtained question queue after removal as the pre-consultation result;

[0113] If the first judgment result is negative, determine the result corresponding to each question in the question queue, and output the result as the pre-consultation result.

[0114] Optionally, the symptom information determination module is specifically configured to:

[0115] Use a large language model to determine whether the description information is related to a preset consultation question, and obtain a second judgment result;

[0116] If the second judgment result is positive, use the large language model to convert the description information to obtain the symptom information of the user to be consulted;

[0117] If the second judgment result is negative, the user to be consulted re-enters the description information.

[0118] Optionally, the question queue construction module is specifically configured to:

[0119] Perform word segmentation on the symptom information of the user to be consulted to obtain a word segmentation result;

[0120] Perform information extraction on the word segmentation result to obtain an information extraction result;

[0121] Based on the information extraction result, determine the symptom attribute questions corresponding to the information extraction result from the knowledge base, and construct a question queue based on the symptom attribute questions; the knowledge base includes symptom attribute questions of various symptoms.

[0122] Optionally, the device further includes: an information acquisition module;

[0123] The information acquisition module is used to acquire the basic information of the user to be consulted; the pre-consultation result corresponds to the basic information.

[0124] Optionally, the device further includes: an upload module;

[0125] The upload module is used to determine whether the user to be consulted uploads a picture, and obtain a third judgment result;

[0126] If the third judgment result is positive, use OCR technology to recognize the picture, obtain a recognition result, and upload the recognition result to the target system;

[0127] If the third judgment result is negative, based on the description information input by the user to be consulted, use a large language model to determine the symptom information of the user to be consulted.

[0128] In addition, an embodiment of the present application further provides a computer-readable storage medium. A computer program is stored in the computer-readable storage medium. When the program is run on a computer device, the pre-consultation method introduced in any manner of the method embodiment is implemented.

[0129] In addition, an embodiment of the present application further provides a computer device. This device is used to run a computer program. When the program runs, it executes the pre-consultation method introduced in any implementation manner of the foregoing method embodiment. As Figure 4 shown, the computer device 01 is presented in the form of a general-purpose computing device. The components of the computer device 01 may include, but are not limited to: one or more processors or processing units 03, a system memory 08, and a bus 04 connecting different system components (including the system memory 08 and the processing unit 03).

[0130] The bus 04 represents one or more of several types of bus structures, including a memory bus or a memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any bus structure in a variety of bus structures. For example, these architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

[0131] The computer device 01 typically includes a variety of computer system-readable media. These media can be any available media that can be accessed by the computer device 01, including volatile and non-volatile media, removable and non-removable media.

[0132] The system memory 08 may include computer system-readable media in the form of volatile memory, such as random access memory (RAM) 09 and / or cache memory 10. The computer device 01 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 11 may be used for reading and writing non-removable, non-volatile magnetic media ( Figure 4 not shown, commonly referred to as a "hard disk drive"). Although Figure 4 not shown in, a disk drive for reading and writing removable non-volatile disks (such as "floppy disks") and an optical disk drive for reading and writing removable non-volatile optical disks (such as CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 04 through one or more data media interfaces. The memory 08 may include at least one program product having a set (for example, at least one) of program modules configured to perform the operations of the embodiments of the present application.

[0133] A program / utilities 12 having a set (at least one) of program modules 13 can be stored, for example, in a memory 08. Such program modules 13 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. The program modules 13 generally execute the functions and / or methods in the embodiments described in the present invention.

[0134] The computer device 01 can also communicate with one or more external devices 02 (such as a keyboard, a pointing device, a display 07, etc.), and can also communicate with one or more devices that enable a user to interact with the computer device 01, and / or communicate with any device that enables the computer device 01 to communicate with one or more other computing devices (such as a network card, a modem, etc.). Such communication can be carried out through an input / output (I / O) interface 06. Moreover, the computer device 01 can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 05. As Figure 4 shown, the network adapter 05 communicates with other modules of the computer device 01 through a bus 04. It should be understood that although Figure 4 not shown in the figure, other hardware and / or software modules can be used in combination with the computer device 01, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0135] The processor unit 03 executes various functional applications and data processing by running programs stored in the system memory 08, such as implementing the pre-consultation method provided in the embodiments of the present application.

[0136] It should be noted that the various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other, and the key points of each embodiment are the differences from other embodiments. In particular, for the device embodiments, since they are basically similar to the method embodiments, they are described relatively simply. For the relevant parts, refer to the description of the method embodiments. The device embodiments described above are only illustrative. The units described as separate components may or may not be physically separated, and the components indicated as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative work.

[0137] As described above, it is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A pre-consultation method, characterized in that, Including: Based on the description information input by the user to be interviewed, use a large language model to determine the symptom information of the user to be interviewed; Use natural language processing technology to extract and process the symptom information of the user to be interviewed, determine the symptom attribute questions corresponding to the extraction and processing results according to the extraction and processing results, and construct a question queue based on the symptom attribute questions; Determine the preliminary interview result of the user to be interviewed based on the extraction and processing results and the question queue.

2. The method according to claim 1, wherein Determine the preliminary interview result of the user to be interviewed based on the extraction and processing results and the question queue, including: Judge whether there is a question corresponding to the extraction and processing result in the question queue, and obtain a first judgment result; If the first judgment result is yes, remove the question corresponding to the extraction and processing result in the question queue to obtain the question queue after removal, and determine the results corresponding to each question in the question queue after removal; Summarize the results corresponding to each question in the obtained question queue after removal according to a preset rule, and output the preliminary interview result; If the first judgment result is no, determine the results corresponding to each question in the question queue, and summarize the obtained results corresponding to each question according to a preset rule, and output the preliminary interview result.

3. The method according to claim 1, wherein Based on the description information input by the user to be interviewed, use a large language model to determine the symptom information of the user to be interviewed, including: Use a large language model to judge whether the description information is related to preset interview questions, and obtain a second judgment result; If the second judgment result is yes, use the large language model to convert the description information to obtain the symptom information of the user to be interviewed; If the second judgment result is no, the user to be interviewed re-enters the description information.

4. The method according to claim 1, wherein Use natural language processing technology to extract and process the symptom information of the user to be interviewed, determine the symptom attribute questions corresponding to the extraction and processing results according to the extraction and processing results, and construct a question queue based on the symptom attribute questions, including: Perform word segmentation processing on the symptom information of the user to be interviewed to obtain a word segmentation processing result; Perform information extraction on the word segmentation processing result to obtain an information extraction result; Based on the information extraction result, determine the symptom attribute questions corresponding to the information extraction result from the knowledge base, and construct a question queue based on the symptom attribute questions; the knowledge base includes symptom attribute questions of various symptoms.

5. The method according to claim 1, wherein Before using a large language model to determine the symptom information of the user to be interviewed based on the description information input by the user to be interviewed, it also includes: Obtain the basic information of the user to be interviewed; the preliminary interview result corresponds to the basic information.

6. The method according to claim 1, characterized in that, Before using a large language model to determine the symptom information of the user to be interviewed based on the description information input by the user to be interviewed, it also includes: Judge whether the user to be interviewed uploads a picture, and obtain a third judgment result; If the third judgment result is yes, use OCR technology to identify the picture to obtain an identification result, and upload the identification result to the target system; If the third judgment result is negative, based on the description information input by the user to be interviewed, use a large language model to determine the symptom information of the user to be interviewed.

7. A pre-consultation device based on a large language model, characterized in that, Including: A symptom information determination module, configured to determine the symptom information of the user to be interviewed by using a large language model based on the description information input by the user to be interviewed; A question queue construction module, configured to perform extraction processing on the symptom information of the user to be interviewed by using natural language processing technology, determine symptom attribute questions corresponding to the extraction processing result based on the extraction processing result, and construct a question queue based on the symptom attribute questions; A preliminary interview result determination module, configured to determine the preliminary interview result of the user to be interviewed based on the processing result and the question queue.

8. The device according to claim 7, wherein The preliminary interview result determination module is specifically configured to: Judge whether there is a question corresponding to the extraction processing result in the question queue, and obtain a first judgment result; If the first judgment result is positive, remove the question corresponding to the extraction processing result in the question queue to obtain a question queue after removal, and determine the result corresponding to each question in the question queue after removal; Summarize the results corresponding to each question in the obtained question queue after removal according to a preset rule, and output a preliminary interview result; If the first judgment result is negative, determine the result corresponding to each question in the question queue, and summarize the results corresponding to each obtained question according to a preset rule, and output a preliminary interview result.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the program is run by a computer device, the preliminary interview method described in any one of claims 1-6 is implemented.

10. A computer device, characterized in that, For running a computer program, when the program runs, it executes the preliminary interview method described in any one of claims 1-6.

Images